> I've yet to see an IDE that isn't horribly slow, bloated and inflexible. I 
> guess in that respect VS is no worse than any other IDE.

Maybe it's your computer then? :-)

VS works perfectly fast enough here thank you... exactly how can an IDE be 
horribly slow? You mean when compiling or starting up or what?

I have Visual C++ 2008 Express here and a moderately complex program (about 
1MB of code in 50 source files using lots of DirectX headers and libraries 
and the Newton Game Dynamics physics library).

Time to open VC++: ~1s
Time to open project: ~1.5s
Time to compile from scratch: ~10s

Post a reply to this message

>> I've yet to see an IDE that isn't horribly slow, bloated and 
>> inflexible. I guess in that respect VS is no worse than any other IDE.
> 
> Maybe it's your computer then? :-)
> 
> VS works perfectly fast enough here thank you... exactly how can an IDE 
> be horribly slow? You mean when compiling or starting up or what?

At uni, we used both Borland JBuilder and VisualStudio J++. Both of them 
were ludicrously slow. Stuff like

- Start the program: 45 seconds.
- Wait for autocomplete to appear: 20 seconds.
- Compile something: 30 seconds.

Both also had a nasty habit of the metadata not matching the source 
code. Like, open a file, add a new method, and VS seems not to "know 
about it" until you close and reopen VS. (You'd hope that just saving 
the file would be enough. No. Maybe just compiling it then? No.)

If you decide you want to rename a class... forget it. This confuses VS 
beyond belief. Trying to add source files that weren't created using VS 
is near impossible. Then VS insists on trying to incorrectly format your 
code (e.g., put the open bracket on the same line as the method 
declaration, rather than the line below it).

Trying to tell it which file is the runnable program is unecessarily 
difficult. You had to jump through hoops to make it create an empty 
project rather than a project already stuffed with boilerplate code you 
don't need or want.

Basically, every single aspect of it was annoying, inflexible and 
overcomplicated.

I will say one thing though: It did manage to compile recursive classes, 
which Sun's JDK could only do with special trickery.

I imagine on today's computers it's probably not quite so slow, but I 
doubt it's become any less inflexible.

Post a reply to this message

>>> 2. Deliberately subverting open standards to force vendor lock-in.
>>
>> That's not illegal, and the vendors were never forced to use the MS 
>> extensions.
> 
> The only one of these I heard of ever being a problem was the Kerberos 
> and/or DHCP standards (I forget which), wherein MS put some important 
> stuff into the area reserved for vendor extensions, which is kind of 
> hard to argue against unless you're fanatical to begin with.

I'm pretty sure that one was Kerberos. And the point was they made it so 
their Kerberos implementation wouldn't work with anybody else's. The 
idea is "if this doesn't work with other compliant Kerberos 
implementations, you can't really call it Kerberos". A bit like if you 
made an SMTP server and an SMTP client that work with each other but not 
any other product on the market; can you really call that SMTP?

If they'd put extra information in but their implementation had still 
worked with other people's, it wouldn't be such a big deal. (It still 
leaves the issue of these fields being reversed for future use, and when 
the next standard decides to use them, it'll break the M$ implementation...)

> What other standards have they done this with.

*cough* The web?

> What I haven't figured out is the huge number of people screaming at 
> Microsoft for bundling programs with Windows that competes with programs 
> they give away for free. Say what?

Before Internet Explorer, web browsers used to cost money. After IE, it 
was basically impossible to sell them for money. Why would anyone pay 
money for something that they can get for free anyway?

So yeah, the products they give away for free compete with other 
products which are free *now*... because you can't sell them for money 
any more.

Post a reply to this message

>> 1. Stealing other people's code and passing it off as their own.
> 
> I've only read of two instances where this was actually proven.  In the 
> first case, it wasn't actually illegal due to poor wording in the 
> license that MS signed with Apple, in the second case it was open code 
> to begin with (the network stack from FreeBSD, I believe).

 From what I heard, Internet Explorer was stolen from a company called 
Spyglass.

Spyglass were selling a web browser, M$ licenced the code from them, 
developed IE, and then gave it away for free (in violation of the terms 
of the licence with Spyglass). Spyglass promptly went bankrupt before 
they could sue M$.

A similar story happened at the beginning of the Microsoft story, but 
since I don't recall the precise facts right now I'll leave that one.

>> 2. Deliberately subverting open standards to force vendor lock-in.
> 
> That's not illegal, and the vendors were never forced to use the MS 
> extensions.

Sure, nobody forces web developers to use IE-specific extensions. Yet 
90% of all websites work properly only if you use IE. (And the others 
only work properly in other browsers because the makers have spent years 
reverse-engineering IE's incorrect behaviour rather than implementing 
the approved standards.)

>> 3. Forcing PC manufacturers to not to distribute anyone else's OS.
> 
> They didn't force anybody.  They signed agreements with manufacturers, 
> and the manufacturers agreed to those contracts.

The way I heard it, it was more like "you will agree to these terms or 
you can't have our product".

>> 4. Lobbying for a ban on selling PC hardware without an OS.
> 
> Name me a big business that doesn't lobby the Government for stuff that 
> will benefit them.  Not only is lobbying legal, but even encouraged by 
> our system.

Yeah, well, when car manufacturers lobby the government saying "people 
aren't buying as many cars as they used to; I think we should get 
government subsidies", people just laugh and say "no". When Microsoft 
lobbies for something, people seem to think they have a point.

>> 5. Releasing software for free just to put people out of business.
> 
> Wait, do you criticize Linus Torvalds for trying to put MS out of 
> business by giving Linux away for free?  And what's the difference 
> between MS selling a better product (and so putting other companies out 
> of business) and MS selling a product for less (or even free) (and so 
> putting other companies out of business)?

Microsoft made their money from Windows. They can afford to give 
products away with it for free. People who's entire business is selling 
those other products can't do this. It's using sales from one product to 
pay for another product; last time I checked, that's not legal.

Like, if Tesco decided to start giving away a free bestselling book with 
every purchase, they'd have a problem because they'd be using grocery 
sales to put book sellers out of business.

>> 6. Announcing vapourware just to put people out of business.
> 
> Sure, they announced it, but you can be assured that if they thought 
> they could make money by selling a particular piece of software they 
> would have made it.  The fact that it ended up being cancelled means 
> they didn't think it would be profitable.

Well, that's fair enough. But when a business announces something they 
have no intention of making just for the financial effect it will have 
on a potential competetor... that's not really fair.

> It sounds like your problem isn't with MS in particular, but with big 
> business in general.

I just dislike the way M$ wins by cheating all the time.

Oracle is a big corporation too, but I have nothing against them. Their 
product is extremely expensive, but it's also high-quality and resonably 
well documented. Oracle didn't get to be where they are now by stealing 
other people's stuff, lying to the government or trying to trick their 
users. They did it by producing a good product.

> Nobody's forcing you to use it.

I'm sure this one has been argued to death. While *technically* this is 
true, the reality is that M$ has carefully engineered a situation where 
little viable alternative actually exists. (Let's face it, if somebody 
else was producing decent software, M$ would go under fairly quickly.)

> Why don't you use Linux instead?

Sure. Except that the vast majority of the software I want to use won't 
work with Linux.

Besides, much as I'd *like* Linux to be the answer... it isn't. It's 
still too old and crufty and messy for my liking. But hey, who's going 
to produce anything better?

Post a reply to this message

> At uni, we used both Borland JBuilder and VisualStudio J++. Both of them 
> were ludicrously slow. Stuff like
>
> - Start the program: 45 seconds.
> - Wait for autocomplete to appear: 20 seconds.

So the computer was to blame then? I used Borland C++ Builder (I got it free 
on a magazine cover disk) extensively at colleg and Uni on my moderate home 
PC and it ran fine, nothing like you mention above.

> If you decide you want to rename a class... forget it. This confuses VS 
> beyond belief.

Did you try to do it manually, or use the Rename command?  If you tried to 
do it manually then I can understand things going wrong.

http://msdn.microsoft.com/en-us/library/6kxxabwd.aspx

> Trying to add source files that weren't created using VS is near 
> impossible.

Ermm, Project -> Add Existing Item works perfectly every time for me.  What 
went wrong?

> Then VS insists on trying to incorrectly format your code (e.g., put the 
> open bracket on the same line as the method declaration, rather than the 
> line below it).

You can do either, and you can choose whether it will indent the brackets or 
not if you put them on their own line.

> Trying to tell it which file is the runnable program is unecessarily 
> difficult.

I've never made any multi-file projects, but I would be very surprised if 
you couldn't set what is run.  And even if you couldn't you just add a 
post-build event to run the executable you want.

> You had to jump through hoops to make it create an empty project rather 
> than a project already stuffed with boilerplate code you don't need or 
> want.

Sounds to me like someone just didn't know how to do it the correct way.

Post a reply to this message

>> At uni, we used both Borland JBuilder and VisualStudio J++. Both of 
>> them were ludicrously slow. Stuff like
>>
>> - Start the program: 45 seconds.
>> - Wait for autocomplete to appear: 20 seconds.
> 
> So the computer was to blame then? I used Borland C++ Builder (I got it 
> free on a magazine cover disk) extensively at colleg and Uni on my 
> moderate home PC and it ran fine, nothing like you mention above.

Well, all the other programs seemed to run just fine. (E.g., starting up 
Word didn't take very long.) I will admit Netscape Navigator always 
seemed to take a long time to start - but it would *run* just fine, it 
just took a while to start.

>> If you decide you want to rename a class... forget it. This confuses 
>> VS beyond belief.
> 
> Did you try to do it manually, or use the Rename command?  If you tried 
> to do it manually then I can understand things going wrong.

IIRC, I used the rename option in VS.

>> Trying to add source files that weren't created using VS is near 
>> impossible.
> 
> Ermm, Project -> Add Existing Item works perfectly every time for me.  
> What went wrong?

Well, first of all, you have to find the project folder on the harddrive 
and dig twenty-five million subfolders down into it to find where the 
actual source code is stored. Then you copy the files there. Then you 
spend half an hour trying to figure out how to make VS realise the files 
exist. And, if you're lucky, it works.

(Recall that Java isn't like C; the source files have to have particular 
names, and have to be located in particular folders, and so on.)

>> Then VS insists on trying to incorrectly format your code (e.g., put 
>> the open bracket on the same line as the method declaration, rather 
>> than the line below it).
> 
> You can do either, and you can choose whether it will indent the 
> brackets or not if you put them on their own line.

I seem to recall spending a long time tweaking the "style" settings of 
the code editor, and I never did get it to work correctly. In the end I 
had to spend lots of time undoing all the autoformatting the editor put 
in just because I couldn't figure out which exact combination of 
settings makes it lay out the code correctly to start with.

>> Trying to tell it which file is the runnable program is unecessarily 
>> difficult.
> 
> I've never made any multi-file projects, but I would be very surprised 
> if you couldn't set what is run.  And even if you couldn't you just add 
> a post-build event to run the executable you want.

Oh, you could set it, it was just unecessarily difficult to do.

(As I stated above, Java *requires* all nontrivial programs to be in 
multiple source files.)

>> You had to jump through hoops to make it create an empty project 
>> rather than a project already stuffed with boilerplate code you don't 
>> need or want.
> 
> Sounds to me like someone just didn't know how to do it the correct way.

Well, it's not like anybody showed us how to use the software...

As far as I could tell, the only thing that VS gives you that Notepad 
doesn't have is the ability to press a button and have it compile 
everything. Sure, it also had that autocomplete thing, but honestly, it 
seemed to get in the way more than actually help. (Aside from being very 
slow and constantly out of date anyway.) You either had to take your 
hand off the keyboard to use the mouse to click the item you wanted, or 
you could select the item with the keyboard and hit return - but then it 
inserted a return into the source. >_<

It did also have a little tree view showing you what files are in your 
project, and one below it showing you the properties of the current 
class - methods, fields, etc. But that seemed to be permanently out of 
date. A nice little feature, but not especially useful.

In summary, I did most of my work with Notepad. It's so much easier to 
get the job done.

Who knows? Perhaps in the 7 years since then, VS has become actually 
useful in some way. But I doubt it...

Post a reply to this message

> Before Internet Explorer, web browsers used to cost money. After IE, it 
> was basically impossible to sell them for money. Why would anyone pay 
> money for something that they can get for free anyway?
>
> So yeah, the products they give away for free compete with other products 
> which are free *now*... because you can't sell them for money any more.

Of course you can if your product is more attractive for customers to want 
to pay for it.  Linux is free, people still buy Windows.  OpenOffice is 
free, people still buy MS Office.  Blender and Wings are free, people still 
buy 3D Studio.  The GIMP is free, people still buy Paintshop.  I'm sure you 
can think of many more examples.

Post a reply to this message

scott wrote:

> Of course you can if your product is more attractive for customers to 
> want to pay for it.

It's difficult to see how you could make a web browser more attractive 
to the casual user than IE already is - at least, enough to make them 
reach into their pockets anyway...

> Linux is free, people still buy Windows.

Some people even buy Linux, oddly enough...

> OpenOffice is free, people still buy MS Office.

A better comparison might be "people still be StarOffice". Except that 
I've never heard of anybody doing that.

> Blender and Wings are 
> free, people still buy 3D Studio.  The GIMP is free, people still buy 
> Paintshop.  I'm sure you can think of many more examples.

If I wanted to be cynical, I could point out that I've never met anybody 
who owns a copy of 3D Studio or Paintshop *legally*. ;-)

Post a reply to this message

> From what I heard, Internet Explorer was stolen from a company called 
> Spyglass.
>
> Spyglass were selling a web browser, M$ licenced the code from them, 
> developed IE, and then gave it away for free (in violation of the terms of 
> the licence with Spyglass). Spyglass promptly went bankrupt before they 
> could sue M$.

Oh joy, another one of your completely made-up stories bashing MS :-)

http://en.wikipedia.org/wiki/Spyglass,_Inc.

> A similar story happened at the beginning of the Microsoft story, but 
> since I don't recall the precise facts right now I'll leave that one.

Yeh, if you are even admitting to not recalling the precise facts this time 
then perhaps best not to :-)

> The way I heard it, it was more like "you will agree to these terms or you 
> can't have our product".

That's impossible, because as pointed out you can just go to a normal shop 
and buy it if you can't get it directly from MS.  The worst MS could do was 
"you will agree to these terms or you will have to buy it at the normal 
price the public pays for it".

> Yeah, well, when car manufacturers lobby the government saying "people 
> aren't buying as many cars as they used to; I think we should get 
> government subsidies", people just laugh and say "no".

Or offer 2000 GBP for your old cars if you buy a new one :-)

> Like, if Tesco decided to start giving away a free bestselling book with 
> every purchase, they'd have a problem because they'd be using grocery 
> sales to put book sellers out of business.

That's not illegal.  In fact it's common practise for supermarkets to use 
"loss leaders" like petrol and milk/sugar that actually lose them money, to 
encourage people to come in and spend on the more profitable items.  What 
*is* illegal is if the manufacturer tries to tell Tesco what price they 
should sell at.

> Well, that's fair enough. But when a business announces something they 
> have no intention of making just for the financial effect it will have on 
> a potential competetor... that's not really fair.

Charities are for being fair, businesses are for making money and crushing 
the competition.

> Oracle is a big corporation too, but I have nothing against them. Their 
> product is extremely expensive, but it's also high-quality and resonably 
> well documented. Oracle didn't get to be where they are now by stealing 
> other people's stuff,

Oh really:

http://en.wikipedia.org/wiki/List_of_acquisitions_by_Oracle

and you should read this:

http://www.zdnetasia.com/news/software/0,39044164,39048963,00.htm

As I said, pretty much any large company is going to have done some "unfair" 
things, otherwise they wouldn't have grown to be a large company.

Post a reply to this message

> (Recall that Java isn't like C; the source files have to have particular 
> names, and have to be located in particular folders, and so on.)

As so the IDE is not really to blame then if you haven't followed the Java 
rules exactly.

> Well, it's not like anybody showed us how to use the software...

LOL.  Reminds me of when you said that 3D Studio was rubbish, but it turns 
out you were just guessing how to use it and got stuck (surprise surprise!).

> Who knows? Perhaps in the 7 years since then, VS has become actually 
> useful in some way. But I doubt it...

Yeh I doubt it too, I guess all those companies than spend tens of thousands 
of dollars on VS licenses are just doing it for fun.  I guess all their 
employees are secretly using notepad...

Post a reply to this message